Athio Intelligence Brief · March 13, 2026

Project Status Map

Paint-by-numbers briefing across all active projects. First principles, second-order effects, gaps, next steps, and how everything connects.

6 Active Projects Last commit: March 13 (Quality Gate) Session resumed after restart
March 13, 2026 — Completed this session:
  • Migrations run: publish_source, swarm_progressive, quality_profile, crm_confidence (sections 19–22)
  • Quality Gate + CRM Confidence committed (11 files) and pushed to Vercel
  • Outreach agents user_id mismatch fixed via WITH clause SQL — agents unblocked
  • This intelligence page published to updates.asapai.net/project-intel-0313

DO THESE FIRST — Priority Queue

1
Deploy Reveal to Vercel (git push)✓ DONE
Quality Gate + CRM Confidence just committed — not live yet. Auto-deploy on push.
2 min
2
Fix Outreach Agents user_id mismatch✓ DONE
1 SQL query in Supabase → 5 agents start running
2 min
3
Test Mastery Lab Playbook API end-to-end
Visit ideas.asapai.net/ai-client-playbook → fill E1 → click Generate → verify report
10 min
4
Reveal build queue: category rename + mode tabs
playbook → lead-magnets (code + DB), then Realtime/Studio mode tabs
~2 hrs
5
Add footer to 4 Mastery Lab pages
Sales Offer, 5-Day Email, Funnel Registry, Internal Registry — add labs.masterymade.com
30 min
6
Publish Kindness validation page
Use Reveal to generate + publish kindness-validation to start market signal collection
30 min
01
Reveal SaaS
AI page generation · nowpage-saas.vercel.app · folio-saas repo
● Live ↑ Just committed March 13
✓
DEPLOYED — March 13, 2026: git push complete. Quality Gate + CRM Confidence now live on Vercel. Auto-deploy triggered.

First Principles

  • Compression engine: unstructured conversation → single URL that proves understanding
  • Trust mechanism: every quote verified against source. No hallucination = credibility.
  • Persona stack = compounding moat: ICP + Voice + Design + CRM accumulate run over run
  • Anti-sales philosophy: the page proves capability. Never pitch. If they can do it without you, say so.
  • The page IS the demo — trust compression in a single artifact

Second-Order Effects

  • Quality gate catches AI-tells before the prospect sees them — credibility compounds over time
  • CRM confidence scoring means the 10th run for a person is dramatically better than the 1st
  • Swarm pipeline (3.9x speedup) enables realtime use during live Zoom calls
  • Each template run teaches the system what "great" looks like for that ICP category
  • Persona stack accumulation = switching cost for power users (their data lives here)
Recent Work (March 5–13)
Mar 13
Quality Gate + CRM Confidence — committed. 4-lens QG (Structural/Persona/Hormozi/Feynman), confidence scoring on CRM/ICP fields. Awaiting Vercel deploy.
Mar 10
Pipeline v3 Swarm — 6 parallel micro-agents, skeleton-first rendering, 3.9x speedup (13.7s vs 53.8s). Circuit breaker for ClawdRouter. JV template fix.
Mar 10
Domain Admin UX — "Subdomain Hosting" rename, delete modal, Archive/Restore, access control fixes.
Mar 9
Quickstarts — saved production configs (domain + voice + design + ICP + template). Wizard UX overhaul: compact pills, config summary, accordion categories, save-as-QS.
Mar 8
MFA + Platform Security — Google Authenticator TOTP, Domain Access v2 (M:N model), admin bypass, unified Users tab.
Mar 7
Live Mode — capture-first flow, realtime transcription, QR code, publish presets (Loop 9.3). ICP/CRM layout redesign.
Mar 6
Custom domain routing fixed — 6 debug commits to resolve 404s. Hub pipeline (triage → hub → sequential). Blueprint injection deployed.
Mar 5
Domain infra — Vercel SSL provisioning CLI/API, ATHIO JV templates (cold/warm/ready blueprints).
Active Build Queue
  1. Category rename: playbook → lead-magnets
    All code + DB update + template seeding. Quick win.
  2. Mode tabs: Realtime / Studio + realtime flag on templates
    PRD: docs/PRD-TEMPLATE-TAXONOMY-AND-MODES.md
  3. New blueprints: sales-pitch, meeting-intel-recap, client-recap
    These are the 3 highest-volume use cases not yet templated
  4. Cost display on template cards (wizard + run detail)
    avg_cost on QS cards, estimated_cost on template cards
  5. Remaining loops: 9.2, 9.4–9.9
    Connected sub-pages, multi-file, structured data, media, deep gates, OAuth, text QR
Key Files & Links
App: /reveal Supabase folio-saas/ (main branch, Vercel auto-deploy) lib/reveal/quality-gate.ts — NEW lib/reveal/crm-confidence.ts — NEW lib/reveal/swarm-agents.ts — Pipeline v3 lib/reveal/blueprints.ts — all templates app/api/reveal/run/route.ts — main pipeline docs/REVEAL-BUILD-PRD.md — 13 layers docs/PRD-TEMPLATE-TAXONOMY-AND-MODES.md
02
Mastery Lab Sales Funnel
ideas.asapai.net · labs.masterymade.com · Feb 25 session
⚠ Mostly Live API Untested
🔴
Middle of funnel broken: app/api/playbook/generate/route.ts is uncommitted and untested. Playbook report generation button may not work in production.

First Principles

  • Qualification-first: only talk to people who are ready (18/27 assessment threshold)
  • Low-friction entry: interactive playbook beats sales call as first touch
  • Email capture as lure: identify signal before investing relationship time
  • Abundant but leave gap: nurture assumes they can do it, but shows why they shouldn't

Second-Order Effects

  • Assessment gate filters time-wasters before any human contact
  • Agentic playbook generates personalized intel before the actual sales call
  • Email nurture warms cold leads so re-engagement is higher quality
  • The funnel itself demonstrates the methodology (using tools to qualify)
  • If Outreach Agents get fixed, they feed directly into this funnel
Page Index
Page Slug Status Blocker
TLDR Registry mastery-lab-tldr ✓ Live —
Agentic Playbook ai-client-playbook ⚠ Live, API untested Test Generate button end-to-end
Assessment mastery-lab-assessment ✓ Live —
Live Session Booking mastery-lab-live-session ⚠ Live, no Calendly Add real Calendly URL
Sales Offer mastery-lab-sales-offer ✗ Missing footer Add labs.masterymade.com footer
5-Day Email Nurture mastery-lab-5day-email-nurture ✗ Missing footer + needs rewrite "Abundant but leave gap" framing
Funnel Registry mastery-lab-funnel-registry ✗ Missing footer Add labs.masterymade.com footer
Internal Registry mastery-lab-internal-registry ✗ Missing footer Add labs.masterymade.com footer
Funnel Flow
ENTRY POINTS ├── AI Client Playbook (free, interactive, email-gated report) └── Live Intelligence Session (assessment-gated booking) │ ▼ QUALIFICATION └── Assessment (9 questions, 27 max, 18 threshold) │ ├── PASS (18+) → Sales Offer (90-day program) → Book Call └── NOT READY (<18) → 5-Day Email Nurture → Reassess
Next Steps
  1. Test playbook API end-to-end
    Visit ideas.asapai.net/ai-client-playbook → fill E1 (2+ fields, 10+ chars each) → click Generate → report should render with scores. Check ANTHROPIC_API_KEY is set in Vercel.
  2. Commit app/api/playbook/generate/route.ts
    File exists locally, type-checks pass, but never pushed to git
  3. Add footer to 4 missing pages
    Add "labs.masterymade.com" footer link to: Sales Offer, 5-Day Email, Funnel Registry, Internal Registry
  4. Rewrite 5-Day Email with "abundant but leave gap" framing
    Current version needs complete reframe — assumes they can do it, but shows why they shouldn't have to
  5. Replace Calendly placeholder on Live Session page
    When booking link is ready, update the mastery-lab-live-session page
03
Outreach Agents
VPS 178.156.253.142 · /opt/outreach-agents/ · Feb 23 session
✓ Unblocked Fixed March 13
✓
FIXED — March 13, 2026: user_id mismatch resolved via SQL. All 5 agents unblocked. Next step: verify signal agent is running on VPS.
THE ONE BLOCKER ✓ DONE
-- APPLIED March 13 — working syntax (subquery in SET failed; WITH clause works): WITH user_lookup AS ( SELECT id FROM auth.users WHERE email = 'jason.macdonald@gmail.com' ) UPDATE domains SET user_id = user_lookup.id FROM user_lookup WHERE domains.user_id IS NULL OR domains.user_id != user_lookup.id;
⚠
Next: SSH to VPS and verify signal agent picked up: journalctl -u outreach-signal --since "5 min ago" — should show [signal] Running... not No config found

First Principles

  • Signal-first: find people expressing pain before approaching them
  • Warm before pitch: 3 touches over 7 days before any direct message
  • Pattern recognition: iterator analyzes rejections to evolve the offer automatically
  • Cost-aware: $0.20–0.68/day total across all 5 agents

Second-Order Effects

  • Self-healing loop: rejection patterns → offer variant evolution without human input
  • Content from Writer agent can be repurposed as lead magnets (flywheel)
  • If it works: ~$3–10/14-day sprint for qualified leads feeding into Mastery Lab funnel
  • Agent architecture is reusable template for any expert's outreach system
Agent Activation Sequence (after fix)
  1. Signal (30min interval)
    Enable first — monitors Reddit/Twitter for pain signals
  2. Scout (4hr interval, after 6h)
    Scores signal authors as ICP prospects
  3. Writer (12hr interval, after 12h)
    Generates content from insights
  4. Warmer (6hr interval, after 24h)
    Engages with prospect content (3 touches over 7 days)
  5. Iterator (24hr interval, after 36h)
    Analyzes rejections, generates offer variants
What's Still Not Done
    • outreach-agents repo not pushed to GitHub — local only at C:/users/jason/downloads/outreach-agents/
    • Signal agent never ran a single successful loop — no baseline data yet
    • Dashboard at /outreach shows 5 agents but no data (agents never ran)
    • Deploy guide (ideas.asapai.net/outreach-deploy-guide-2026-02-22) references old structure
04
Kindness Platform
Richard Edlich · C:/users/jason/downloads/rich/ · Validation phase
🔵 Validation Phase Pages Built

First Principles

  • Lowest-friction capture: voice input beats any form. Reduce to one utterance.
  • Self-actualization audience: people already inclined toward kindness — don't motivate, just count
  • Seed and watch: plant minimal experiments, nurture only what sprouts
  • Social proof loop: global counter creates belonging (22M doors held open globally)

Second-Order Effects

  • Church integration = captive audience pre-qualified by values (high conversion)
  • Dinner table prompt = family ritual = daily habit formation without marketing spend
  • Global counter creates network effect — more users = more social proof = more users
  • If validated → "Strava for kindness" (public leaderboard, team challenges, corporate CSR)
  • Donation loop-back mechanism = monetization that reinforces the mission
Pages Built (in rich/ folder)
kindness-playbook.html kindness-validation.html kindness-dash-v2.html kindness-ai-demo-v2.html meeting-intel-v2.html stories-v2.html bridge-the-gap.html
⚠
Gap: No clear timeline or commitment from Rich. Need to define what "validated" means before building the product. Recommend: pick ONE validation lure, set a 2-week test window, define the signal threshold.
Next Steps
  1. Define validation criteria with Rich
    What signal = "we should build"? Email signups? App downloads? Church pilot response rate?
  2. Publish kindness-validation.html as the lure
    Deploy to ideas.asapai.net or dedicated domain. Use email capture as signal.
  3. Identify one church for pilot
    Pre-service prompt: "This week, hold a door for someone. That's it. We'll count it." → QR code
  4. Build minimal voice prototype
    Browser SpeechRecognition API + NowPage backend. "Hey, I held a door." → +1 counter.
  5. 2-week pilot → review signal
    If >50 unprompted entries → validated. Build the next layer.
05
Tony Gaskins Pages
Rich's client · Built March 9 · Awaiting publish decision
Built — Not Published

Files Built

  • tony-gaskins-meeting-intel-v2.html — meeting intelligence page from Rich's session
  • tony-gaskins-playbook-cold-v2.html — cold outreach playbook page

Next Steps

  • Review both pages with Rich — confirm they're ready
  • Decide on target domain + slug
  • Publish via hc-publish.js or Reveal dashboard
  • Get feedback from Rich / Tony on the output
06
Jason MacDonald Personal Page
Built March 8–9 · 3 versions exist
V3 Latest

File Versions

  • jason-macdonald-v1.html — Built March 8 (10:51 PM)
  • jason-macdonald-v2.html — Built March 9 (11:54 AM)
  • jason-macdonald-v3.html — Built March 9 (12:49 PM) — latest
⚠
Confirm which version is live vs. which is the working draft. If v3 is ready, publish to clean slug and archive v1/v2 per versioning convention.
↔
Cross-Project Alignment
How the 6 projects interconnect and reinforce each other
Outreach → Mastery Lab
→ Agents find pain signals on Reddit/Twitter
→ Prospects enter the Mastery Lab funnel
→ Assessment gates unqualified leads
STALLED: user_id mismatch blocks this flow
Reveal → Rich's Pages
→ Rich's transcripts feed into Reveal
→ Reveal generates Tony Gaskins + Kindness pages
→ Quality Gate now checks these before publish
LIVE: Quality Gate just deployed
Reveal → Mastery Lab
→ Mastery Lab funnel pages could be Reveal-generated
→ Agentic Playbook is Reveal thinking applied to funnel
→ Same anti-sales, page-as-proof philosophy
Kindness → Reveal
→ Kindness validation page = Reveal use case demo
→ Church pilot pages generated by Reveal
→ Meta: using the tool to build the tool's use cases
Outreach → Reveal
→ Writer agent insights → Reveal lead magnet pages
→ Prospect intel → CRM profile → better Reveal pages
→ Content engine flywheel: signal → page → lead
Quality Gate → Everything
→ All future Reveal pages pass 4-lens check
→ Tony Gaskins / Kindness pages now get QG'd
→ Mastery Lab pages could be retroactively audited
DEPLOYED ✓ March 13
⚠
Gaps & Risks
Things that are broken, unverified, or could compound into bigger problems

🔴 Immediate Risks

  • Reveal not deployed: Quality Gate committed but Vercel hasn't picked it up yet — push required ✓ DEPLOYED
  • Outreach agents user_id mismatch: SQL fix applied March 13 ✓ FIXED
  • Outreach agents: first run pending. Fix applied — verify signal agent actually fires on VPS before marking resolved.
  • Mastery Lab API untested: The middle of the funnel (playbook report) may be broken in prod
  • Calendly placeholder: Live Session booking page has no real URL — dead end in funnel

⚠ Medium-Term Gaps

  • ClawdRouter still down: Circuit breaker now handles it gracefully, but primary LLM route is off
  • outreach-agents not on GitHub: If VPS goes down, code is local-only
  • Kindness timeline unclear: No commitment from Rich on what "validated" looks like
  • Swarm orchestrator migration: 20260309_swarm_orchestrator.sql is untracked — may not be applied
  • MEMORY.md shows "Code Complete — Needs Deploy" for Audit Layer + Multi-Page Blueprints — verify these actually deployed

📋 Alignment Gaps

  • Outreach feeds Mastery Lab but is stalled — the supply chain for new leads is broken
  • Kindness Platform has no definition of success — risk of building without validating
  • Jason MacDonald page: 3 versions, unclear which is the canonical live version
  • Tony Gaskins pages built but not published — no feedback loop from the prospect
  • Rich has transcripts → we have pages → but are the pages actually being used/shared?
Existential Risk (Reveal SaaS)

⚡ Primary Risk

  • LLM providers ship native "meeting to artifact" feature (Anthropic, OpenAI, Zoom AI)
  • Defense: Persona depth + verification + blueprint craft. Code is clonable; taste is not.
  • CRM Confidence scoring (just shipped) is a key moat element — gets better with every run
  • Quality Gate (just shipped) is a key moat element — catches what LLM providers miss
  • Quickstarts + Blueprints = institutional knowledge that compounds. Can't clone a user's 50 runs of context.